Comment by thaumasiotes | Hacker News Reader

Comment by thaumasiotes | Hacker News Reader

Parent Full thread

> Select the English phrase that correctly translates the Latin.

> Fīliae aquam agricolae dant.

> 1. The daughters give the farmer water.

> 2. The farmers give (their) daughters water.

> 3. The farmer gives (his) daughters water.

> 4. The farmer gives (his) daughter water.

This is answerable, but only because three of the answers are impossible. You have to make the assumption that the correct translation is given as one of the options. Unmentioned but possible readings include "The farmers give water to the daughter" and "The farmer's daughters give water [to something determined by context]".

If I'm going to be asked to do translation exercises, I'd rather have questions that I need to answer by understanding the text, as opposed to questions that I need to answer by referring to the construction of the question. Follow that sentence with something like Aqua bibita, puellis gratias agit and suddenly we can determine what the first sentence meant regardless of which options we're given.

> Choose the word that matches the indicated use.

> Genitive of the possesor

> 1. corōnam and Diānam in "Puella Diānae corōnam dat quia Diānam amat."

> 2. dominīs in "Nautae fābulam dominīs nārrat quia dominae fābulās amant."

> 3. Cui in "Cui puellae corōnam dant?"

This is much worse; these three sentences between them do not contain even one word in the genitive case. Also, there's a typo in #2, which should presumably say either "Nautae... narrant" or "Nauta... narrat". "Possesor" is a typo too.

I picked #2 and was told it was correct. That grayed the question out, so I can't tell whether options 1 or 3 would also have been "correct".

> Genitive of the possesor

> 3. pugnae in "Nūntiat Galba nautae causam pugnae."

This one is at least genitive, but I have trouble characterizing this as "genitive of the possessor". Going by the taxonomy here ( https://classics.osu.edu/Undergraduate-Studies/Latin-Program... ), I'd call it an objective genitive.

That particular sentence is a really tough knot pedagogically too. As a teacher, you want a short simple sentence that includes a nominative, genitive, dative, accusative, and verb to habituate students to thinking about cases.

Unfortunately, in the first declension, the nominative plural, genitive singular, and and dative singular all look identical, hence the confusion here. Filiae and agricolae could be any of those three options, which means, as you rightly say, that we need context to render the sentence. That introduced students to another distinctive feature of Latin: like C++, Latin has a context sensitive grammar.

If you wanted to separate the introduction of the case system from the introduction of the context sensitivity of the grammar, you could do this sentence:

> Filiabus aquam agricolae dant.

While that fixes most of the context sensitivity, it doesn't fix all of it, and introduces an irregular declension, also tough right out the gate.

Latin is a hard language to teach.

Actually this is an example of ambiguity, not context sensitivity (in the formal languages sense). Context sensitivity means context is needed to construct a grammatical sentence, not merely to parse one. It's actually pretty hard to find examples of this in natural language.

Judging by the paper linked sidethread, it's enough that some valid sentences cannot be constructed in a context-free way. Even if they can be losslessly transformed into alternative sentences which can be easily derived from a context-free grammar, their existence demonstrates that the language is not context-free.

> That introduced students to another distinctive feature of Latin: like C++, Latin has a context sensitive grammar.

Surely this is a feature of all languages :)

Certainly not.

The first example of context-sensitive constructions in natural language was found in 1985.

http://www.eecs.harvard.edu/~shieber/Biblio/Papers/shieber85...

Looking at your second example...that's just a mistake, sorry. I figured a couple of those were bound to slip in. I'll get it fixed. And again, I really appreciate your taking a look at the site.

Thanks for the feedback! As you noticed, the drills are more about syntax than reading comprehension, and your understanding of syntax is solid. It might be challenging to come up with multiple correct translations of each Latin exercise, but I will give some thought to how it might be done.

One way I've done that in the past is asking explicit questions about the syntax, like:

> "If filiae is nominative, how would you translate the sentence?"

or

> What cases could agricolae be here?

if you want to introduce the context sensitivity of the grammar.

Those are two great ideas for question formats. I will likely try to work those into future quizzes. Thanks for checking out the site!

Sure! It's good stuff.

#2 you could read, "[He] tells the sailor's story to the mistresses, because mistresses love stories", which does have a possessive genitive.

Fair enough. I might have thought more about that if nautae had been one of the options for "which word is genitive?"

Yes, I think that was the original intention. I changed the option to - Nautae in "Nautae fābulam dominīs nārrat quia dominae fābulās amant." - for the time being, so at least it isnt't flat out wrong while I think of something better.

Reply on news.ycombinator.com